AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Effect of Prompt Optimization on Pairwise Causal Discovery
We didn't do much exploration in terms of prompt optimization, right? So we just basically gave it a kind of obvious system prompt that stood out to us as for chain of thought reasoning. There was a clear sign that better prompting would have an impact, but that was not the dimension that we explored deeply. It did very well in pairwise causal discovery; there's some concern about the validity of that as a result. We want it to know what altitude is and what temperature is and have memorized those, you know, it's not an agent acting in the world.